Despite the fact that little is known about non-genetic causes of autism spectrum disorders (ASD), currently, few rigorous investigations of environmental factors in the etiology of this condition are underway. This proposal extends the epidemiology project initiated under the NIEHS-funded UC Davis Center for Children's Environmental Health (CCEH) known as MARBLES (Markers of Autism Risk in Babies - Learning Early Signs). MARBLES has enrolled close to 200 pregnant women who already have a child with ASD and therefore are at high risk for delivering an infant who will also develop ASD. In this R01, we will recruit an additional 250 pregnant mothers and follow their pregnancy and their child. Through longitudinal collection of 1) extensive behavioral, medical, and exposure data, 2) biologic specimens from the mother and child, and 3) detailed psychometric assessments from birth to three years of age, MARBLES has created an infrastructure for addressing etiologic questions and searching for early pathophysiologic markers. The exposures of interest are two classes of compounds common in household products, and having known neuro- or neurodevelopmental toxicity: pyrethroid pesticides and polybrominated diphenyl ethers (flame retardants). First, we will assess associations of self-reported exposure, measurements of internal dose, and toxicologically-derived estimates of biologically effective dose, on the one hand, with risk for ASD or other impairments in neurobehavioral development on the other. Secondly, we will examine whether the exposure or dose estimates are associated with markers of aberrant immune responses or mitochondrial dysfunction and whether these markers predict clinically confirmed child developmental status at three years of age. Thus, this project begins with the macro-level associations typical of black-box epidemiology, refines the estimates of dose, and then explores more deeply into the mechanisms by which environmental exposures might alter neurodevelopment and lead to clinical outcomes. This integrated approach is made possible by an interdisciplinary team that brings together molecular and basic research, epidemiologic population-based approaches, and clinical sciences. If our hypotheses are supported, this will be the first evidence of higher autism risk based on prospective measurements of compounds in commonly used household products, and one of the first to identify modifiable risk factors. The results will thereby open the door to prevention at the individual behavioral and societal level, e.g., potentially through education, product labeling, and/or regulatory action. Results may also lead to targeted interventions to alter children's developmental trajectory. Overall, we expect this study to stimulate a paradigm shift in the field, towards a more multifactorial and mechanistically driven research agenda, with significantly more attention to environmental exposures and the development of inter-disciplinary approaches.